Prediction of Admission Costs Following Anterior Cervical Discectomy and Fusion Utilizing Machine Learning.

STUDY DESIGN: Retrospective case series. OBJECTIVE: Predict cost following anterior cervical discectomy and fusion (ACDF) within the 90-day global period using machine learning models. BACKGROUND: The incidence of ACDF has been increasing with a disproportionate decrease in reimbursement. As bundled payment models become common, it is imperative to identify factors that impact the cost of care. MATERIALS AND METHODS: The Nationwide Readmissions Database (NRD) was accessed in 2018 for all primary ACDFs by the International Classification of Diseases 10th Revision (ICD-10) procedure codes. Costs were calculated by utilizing the total hospital charge and each hospital's cost-to-charge ratio. Hospital characteristics, such as volume of procedures performed and wage index, were also queried. Readmissions within 90 days were identified, and cost of readmissions was added to the total admission cost to represent the 90-day healthcare cost. Machine learning algorithms were used to predict patients with 90-day admission costs >1 SD from the mean. RESULTS: There were 42,485 procedures included in this investigation with an average age of 57.7±12.3 years with 50.6% males. The average cost of the operative admission was $24,874±25,610, the average cost of readmission was $25,371±11,476, and the average total cost was $26,977±28,947 including readmissions costs. There were 10,624 patients who were categorized as high cost. Wage index, hospital volume, age, and diagnosis-related group severity were most correlated with the total cost of care. Gradient boosting trees algorithm was most predictive of the total cost of care (area under the curve=0.86). CONCLUSIONS: Bundled payment models utilize wage index and diagnosis-related groups to determine reimbursement of ACDF. However, machine learning algorithms identified additional variables, such as hospital volume, readmission, and patient age, that are also important for determining the cost of care. Machine learning can improve cost-effectiveness and reduce the financial burden placed upon physicians and hospitals by implementing patient-specific reimbursement.

Prediction of total healthcare cost following total shoulder arthroplasty utilizing machine learning.

BACKGROUND: Given the increase in demand in treatment of glenohumeral arthritis with anatomic total (aTSA) and reverse shoulder arthroplasty (RTSA), it is imperative to improve quality of patient care while controlling costs as private and federal insurers continue its gradual transition toward bundled payment models. Big data analytics with machine learning shows promise in predicting health care costs. This is significant as cost prediction may help control cost by enabling health care systems to appropriately allocate resources that help mitigate the cause of increased cost. METHODS: The Nationwide Readmissions Database (NRD) was accessed in 2018. The database was queried for all primary aTSA and RTSA by International Classification of Diseases, Tenth Revision (ICD-10) procedure codes: 0RRJ0JZ and 0RRK0JZ for aTSA and 0RRK00Z and 0RRJ00Z for RTSA. Procedures were categorized by diagnoses: osteoarthritis (OA), rheumatoid arthritis (RA), avascular necrosis (AVN), fracture, and rotator cuff arthropathy (RCA). Costs were calculated by utilizing the total hospital charge and each hospital's cost-to-charge ratio. Hospital characteristics were included, such as volume of procedures performed by the respective hospital for the calendar year and wage index, which represents the relative average hospital wage for the respective geographic area. Unplanned readmissions within 90 days were calculated using unique patient identifiers, and cost of readmissions was added to the total admission cost to represent the short-term perioperative health care cost. Machine learning algorithms were used to predict patients with immediate postoperative admission costs greater than 1 standard deviation from the mean, and readmissions. RESULTS: A total of 49,354 patients were isolated for analysis, with an average patient age of 69.9 ± 9.6 years. The average perioperative cost of care was $18,843 ± $10,165. In total, there were 4279 all-cause readmissions, resulting in an average cost of $13,871.00 ± $14,301.06 per readmission. Wage index, hospital volume, patient age, readmissions, and diagnosis-related group severity were the factors most correlated with the total cost of care. The logistic regression and random forest algorithms were equivalent in predicting the total cost of care (area under the receiver operating characteristic curve = 0.83). CONCLUSION: After shoulder arthroplasty, there is significant variability in cumulative hospital costs, and this is largely affected by readmissions. Hospital characteristics, such as geographic area and volume, are key determinants of overall health care cost. When accounting for this, machine learning algorithms may predict cases with high likelihood of increased resource utilization and/or readmission.

Liposomal Bupivacaine Associated with Cost Savings during Postoperative Pain Management in Fragility Intertrochanteric Hip Fractures.

Background: Intertrochanteric hip fractures are among the most common and most expensive diagnoses in the Medicare population. Liposomal bupivacaine is a novel preparation of a commonly used analgesic agent that, when used intraoperatively, decreases narcotic requirements and hospital length of stay and increases the likelihood of discharge to home. The purpose of this investigation was to determine whether there was an economic benefit to utilizing intraoperative liposomal bupivacaine in patients with fragility intertrochanteric hip fractures in comparison to a group of patients who did not receive liposomal bupivacaine. Methods: This is a retrospective observational study performed at two academic medical centers. Fifty-six patients with intertrochanteric hip fractures treated with cephalomedullary nail implant who received standard hip fracture pain management protocol were compared to a cohort of 46 patients with intertrochanteric hip fractures who received additional intraoperative injections of liposomal bupivacaine. All other standards of care were identical. A cost analysis was completed including the cost of liposomal bupivacaine, operating room costs, and discharge destination. Statistical significance was set at p < 0.05. Results: Although the length of hospital stay was similar between the two groups (3.2 days vs. 3.8 days, p = 0.08), patients receiving intraoperative liposomal bupivacaine had a lower likelihood of discharge to a skilled nursing facility (84.8% vs. 96.4%, p = 0.002) and a longer operative time (73.4 minutes vs 67.2 minutes, p = 0.004). The cost-benefit analysis indicated that for an investment of $334.18 in the administration of 266 mg of liposomal bupivacaine, there was a relative saving of $1,323.21 compared to the control group. The benefit-cost ratio was 3.95, indicating a $3.95 benefit for each $1 spent in liposomal bupivacaine. Conclusions: Despite the increased initial cost, intraoperative use of liposomal bupivacaine was found to be a cost-effective intervention due to the higher likelihood of discharge to home during the postoperative management of patients with intertrochanteric hip fractures.

A Risk Stratification Nomogram to Predict Inpatient Admissions After Total Shoulder Arthroplasty Among Patients Eligible for Medicare.

The goal of this study was to establish a risk stratification nomogram to aid in determining the need for inpatient admission among patients who were eligible for Medicare and were undergoing primary total shoulder arthroplasty (TSA). The American College of Surgeons National Surgical Quality Improvement Program database was queried to identify all patients older than 65 years who underwent primary TSA between 2006 and 2016. The primary outcome measure was inpatient admission, as defined by hospital length of stay longer than 2 days. Multiple demographic, comorbid, and peri-operative variables were used in a multivariate logistic regression model to yield a risk stratification nomogram. A total of 1514 inpatient and 6020 out-patient admissions were analyzed. Age older than 80 years (odds ratio [OR], 2.69; P<.0001; 95% CI, 2.21-3.27), female sex (OR, 2.18; P<.0001; 95% CI, 1.90-2.51), dependent functional status (OR, 1.69; P<.0001; 95% CI, 1.2-2.38), dialysis (OR, 3.48; P=.029; 95% CI, 1.14-10.63), admission from an inpatient facility (OR, 1.76; P<.0001; 95% CI, 1.70-1.82), and inflammatory arthritis (OR, 1.69; P<.02; 95% CI, 1.25-13.78) were the greatest determinants of inpatient stay. The resulting predictive model showed acceptable discrimination and calibration. Our model enabled reliable and straightforward identification of the most suitable candidates for inpatient admission among patients who were eligible for Medicare and were undergoing primary TSA. Patients who were receiving dialysis, who had dyspnea at rest, and who had bleeding disorders were more likely to be admitted as inpatients after TSA. Larger multicenter studies are necessary to externally validate the proposed predictive nomogram. [Orthopedics. 2022;45(1):43-49.].

Predicting Patient Satisfaction With Maximal Outcome Improvement After Biceps Tenodesis.

The goal of this study was to determine the threshold for achieving maximal outcome improvement (MOI) on the American Shoulder and Elbow Surgeons (ASES), Single Assessment Numeric Evaluation (SANE), and Constant-Murley (CM) questionnaires that predict satisfaction after isolated biceps tenodesis without concomitant rotator cuff repair. A retrospective analysis of prospectively collected data was performed for patients undergoing isolated biceps tenodesis from 2014 to 2017 at a single institution with minimum 6-month follow-up. Receiver operating characteristic curve analysis was used to determine thresholds for MOI for the ASES, SANE, and CM questionnaires. Stepwise multivariate logistical regression analysis was performed to identify predictors for achieving the threshold for MOI. A total of 123 patients were included in the final analysis. Receiver operating characteristic analysis determined that achieving 43.1%, 62.1%, and 61.4% MOI was the threshold for satisfaction for the ASES, SANE, and CM questionnaires, respectively. Regression analysis showed that concomitant superior labrum anterior-posterior (SLAP) repair was predictive of achieving MOI on the ASES and SANE questionnaires, whereas partial rotator cuff tear was predictive of achieving MOI on the CM questionnaire (P<.05 for both). Further, workers' compensation status, diabetes, history of ipsilateral shoulder surgery, and hypertension were negative predictors of achieving MOI on the SANE and CM questionnaires (P<.05 for all). Achieving MOI of 43.1%, 62.1%, and 61.4% is the threshold for satisfaction after biceps tenodesis for the ASES, SANE, and CM questionnaires, respectively. Concomitant SLAP repair was positively predictive of achieving MOI, whereas workers' compensation status, diabetes, history of ipsilateral shoulder surgery, and hypertension were negative predictors. [Orthopedics. 2021;44(3):e359-e366.].

Assessment and Trends of the Methodological Quality of the Top 50 Most Cited Articles on Patellar Instability.

BACKGROUND: Studies with a low level of evidence (LOE) have dominated the top cited research in many areas of orthopaedics. The wide range of treatment options for patellar instability necessitates an investigation to determine the types of studies that drive clinical practice. PURPOSE: To determine (1) the top 50 most cited articles on patellar instability and (2) the correlation between the number of citations and LOE or methodological quality. STUDY DESIGN: Cross-sectional study. METHODS: The Scopus and Web of Science databases were assessed to determine the top 50 most cited articles on patellar instability between 1985 and 2019. Bibliographic information, number of citations, and LOE were collected. Methodological quality was calculated using the Modified Coleman Methodology Score (MCMS) and the Methodological Index for Non-Randomized Studies (MINORS). Mean citations and mean citation density (citations per year) were correlated with LOE, MCMS, and MINORS scores. RESULTS: Most studies were cadaveric (n = 10; 20.0%), published in the American Journal of Sports Medicine (n = 13; 26.0%), published between 2000 and 2009 (n = 41; 82.0%), and conducted in the United States (n = 17; 34.0%). The mean number of citations and the citation density were 158.61 ± 59.53 (range, 95.5-400.5) and 12.74 ± 5.12, respectively. The mean MCMS and MINORS scores were 59.62 ± 12.58 and 16.24 ± 3.72, respectively. No correlation was seen between mean number of citations or citation density versus LOE. A significant difference was found in the mean LOE of articles published between 1990 and 1999 (5.0 ± 0) versus those published between 2000 and 2009 (3.12 ± 1.38; P = .03) and between 2010 and 2019 (3.00 ± 1.10; P = .01). CONCLUSION: There was a shift in research from anatomy toward outcomes in patellar instability; however, these articles demonstrated low LOE and methodological quality. Higher quality studies are necessary to establish informed standards of management of patellar instability.

Return to work following distal triceps repair.

PURPOSE: The purpose of this study was to evaluate the rate and duration of return to work in patients undergoing distal triceps repair (DTR). METHODS: Consecutive patients undergoing DTR from 2009 to 2017 at our institution were retrospectively reviewed at a minimum of 1 year postoperatively. Patients completed a standardized and validated work questionnaire; a visual analog scale for pain; the Mayo Elbow Performance Score; the short version of the Disabilities of the Arm, Shoulder and Hand questionnaire; and a satisfaction survey. RESULTS: Of 113 eligible patients who underwent DTR, 81 (71.7%) were contacted. Of these patients, 74 (91.4%) were employed within 3 years prior to surgery (mean age, 46.0 ± 10.7 years; mean follow-up, 5.9 ± 3.9 years). Sixty-nine patients (93.2%) returned to work by 2.2 ± 3.2 months postoperatively. Sixty-six patients (89.2%) were able to return to the same level of occupational intensity. Patients who held sedentary-, light-, medium-, and high-intensity occupations were able to return to work at a rate of 100.0%, 100.0%, 80.0%, and 76.9%, respectively, by 0.3 ± 0.5 months, 1.8 ± 1.5 months, 2.5 ± 3.6 months, and 4.8 ± 3.9 months, respectively, postoperatively. Of the workers' compensation patients, 15 (75%) returned to work by 6.5 ± 4.3 months postoperatively, whereas 100% of non-workers' compensation patients returned to work by 1.1 ± 1.6 months (P < .001). Seventy-one patients (95.9%) were at least somewhat satisfied, with 50 patients (67.6%) reporting excellent satisfaction. Seventy-two patients (97.3%) would undergo the operation again if presented the opportunity. A single patient (1.4%) required revision DTR. CONCLUSIONS: Approximately 93% of patients who underwent DTR returned to work by 2.2 ± 3.2 months postoperatively. Patients with higher-intensity occupations had an equivalent rate of return to work but took longer to return to their preoperative level of occupational intensity. Information regarding return to work is imperative in preoperative patient consultation to manage expectations.

Influence of workers' compensation status on postoperative outcomes in patients following biceps tenodesis: a matched-pair cohort analysis.

BACKGROUND AND HYPOTHESIS: Although the literature on the association of workers' compensation (WC) status with negative outcomes after orthopedic surgery is extensive, there is a paucity of evidence on outcomes in WC recipients undergoing biceps tenodesis. We hypothesized that WC patients would report significantly worse outcomes postoperatively on patient-reported outcome measures (PROMs). METHODS: Functional and health-related quality-of-life PROMs and a visual analog scale score for pain were administered preoperatively and at 12 months postoperatively to consecutive patients undergoing isolated biceps tenodesis between 2014 and 2018 at our institution. Thirty-eight WC patients were matched 1:2 to non-WC patients by age, body mass index, and operative limb. The minimal clinically important difference, substantial clinical benefit, and patient acceptable symptom state were calculated for all patients via anchor- and distribution-based methods. Rates of achievement and the likelihood of achievement were determined. RESULTS: All patients showed significant improvements in all outcome measures (P < .001). WC patients reported inferior postoperative scores on all PROMs examined. WC status significantly predicted a reduced likelihood of achieving substantial clinical benefit for the American Shoulder and Elbow Surgeons score (odds ratio [OR], 0.38; 95% confidence interval [CI], 0.17-0.81; P = .01) and the patient acceptable symptom state (OR, 0.28; 95% CI, 0.12-0.65; P = .003) for the American Shoulder and Elbow Surgeons score, Single Assessment Numeric Evaluation score (OR, 0.24; 95% CI, 0.10-0.61; P = .003), Constant-Murley Subjective Assessment (OR, 0.25; 95% CI, 0.08-0.77; P = .016), and visual analog scale pain score (OR, 0.27; 95% CI, 0.16-0.47; P < .001). CONCLUSION: WC patients reported inferior scores on all postoperative PROMs and demonstrated lower odds of achieving substantial benefit and satisfaction regarding improvements in both function and pain compared with non-WC patients.

Patient Satisfaction After Total Shoulder Arthroplasty.

Although patient-reported outcome measures use objective evaluations of impairment to focus on subjective responses, these measures may not necessarily reflect patient satisfaction with the outcome or the care provided. The goal of this study was to systematically review the available literature to assess patient satisfaction after total shoulder arthroplasty. Two investigators systematically reviewed the MEDLINE database for articles on satisfaction after this procedure. This study included 47 articles. The most commonly used method for assessing satisfaction was an ordinal scale (27 studies, 57.4%). Of the studies, 27 (57.5%) differentiated between patient satisfaction with the care provided and with the outcome achieved. Reported satisfaction rates after anatomic total shoulder arthroplasty ranged from 75% to 100%. For the included studies, increasing age, workers' compensation status, depression, opioid use, and visual analog scale pain score were the only preoperative factors that were significantly associated with worse postoperative satisfaction. Postoperative American Shoulder and Elbow Surgeons score, Simple Shoulder Test score, Subjective Shoulder Value score, Short Form-36 mental component score, range of motion, visual analog scale pain score, and ability to perform activities of daily living showed a significant association with postoperative satisfaction. Studies of satisfaction after total shoulder arthroplasty are of low evidence levels. Although overall patient satisfaction is high, there is no standardized method for measuring satisfaction. For the identified studies, the most common assessment method was an ordinal scale that consists of qualitative values representing increasing levels of satisfaction. Orthopedic surgeons are increasingly expected to demonstrate the value of procedures, and a uniform and validated method of assessing patient satisfaction is needed. [Orthopedics. 2020;43(6):e492-e497.].

Big Data in Total Shoulder Arthroplasty: An In-depth Comparison of National Outcomes Databases.

INTRODUCTION: The practice of identifying trends in surgical decision-making through large-scale patient databases is commonplace. We hypothesize that notable differences exist between claims-based and prospectively collected clinical registries. METHODS: We queried the American College of Surgeons National Surgical Quality Improvement Program (NSQIP), a prospective surgical outcomes database, and PearlDiver (PD), a claims-based private insurance database, for patients undergoing primary total shoulder arthroplasties from 2007 to 2016. Comorbidities and 30-day complications were compared. Multiple regression analysis was performed for each cohort to identify notable contributors to 30-day revision surgery. RESULTS: Significant differences were observed in demographics, comorbidities, and postoperative complications for the age-matched groups between PD and NSQIP (P < 0.05 for all). Multiple regression analysis in PD identified morbid obesity and dyspnea to lead to an increased risk for revision surgery (P = 0.001) in the <65 cohort and dyspnea and diabetes to lead to an increased risk for revision surgery in the ≥65 cohort (P = 0.015, P < 0.001). Multiple regression did not reveal any risk factors for revision surgery in the <65 age group for the NSQIP; however, congestive heart failure was found to have an increased risk for revision surgery in the ≥65 cohort (P < 0.001). CONCLUSIONS: Notable differences in comorbidities and complications for patients undergoing primary total shoulder arthroplasty were present between PD and NSQIP. LEVEL OF EVIDENCE: Retrospective cohort study, level III.

Assessment and Trends in the Methodological Quality of the Top 50 Most Cited Articles in Shoulder Instability.

BACKGROUND: Citation counts have often been used as a surrogate for the scholarly impact of a particular study, but they do not necessarily correlate with higher-quality investigations. In recent decades, much of the literature regarding shoulder instability is focused on surgical techniques to correct bone loss and prevent recurrence. PURPOSE: To determine (1) the top 50 most cited articles in shoulder instability and (2) if there is a correlation between the number of citations and level of evidence or methodological quality. STUDY DESIGN: Cross-sectional study. METHODS: A literature search was performed on both the Scopus and the Web of Science databases to determine the top 50 most cited articles in shoulder instability between 1985 and 2019. The search terms used included "shoulder instability," "humeral defect," and "glenoid bone loss." Methodological scores were calculated using the Modified Coleman Methodology Score (MCMS), Jadad scale, and Methodological Index for Non-Randomized Studies (MINORS) score. RESULTS: The mean number of citations and mean citation density were 222.7 ± 123.5 (range, 124-881.5) and 16.0 ± 7.9 (range, 6.9-49.0), respectively. The most common type of study represented was the retrospective case series (evidence level, 4; n = 16; 32%) The overall mean MCMS, Jadad score, and MINORS score were 61.1 ± 10.1, 1.4 ± 0.9, and 16.0 ± 3.0, respectively. There were also no correlations found between mean citations or citation density versus each of the methodological quality scores. CONCLUSION: The list of top 50 most cited articles in shoulder instability comprised studies with low-level evidence and low methodological quality. Higher-quality study methodology does not appear to be a significant factor in whether studies are frequently cited in the literature.

Establishing minimal clinically important difference, substantial clinical benefit, and patient acceptable symptomatic state after biceps tenodesis.

BACKGROUND: The purposes of this study were to establish thresholds for improvement in patient-reported outcome scores that signify the minimal clinically important difference (MCID), substantial clinical benefit (SCB), and patient acceptable symptomatic state (PASS) after biceps tenodesis (BT) and to assess patient variables that are associated with these clinically important outcomes. METHODS: A prospectively maintained institutional shoulder registry was queried for patients undergoing isolated BT between 2014 and 2017. Anchor-based and distribution-based approaches were used to calculate the MCID whereas an anchor-based method was used to calculate the SCB and PASS for the Constant-Murley score, Single Assessment Numerical Evaluation (SANE) score, and American Shoulder and Elbow Surgeons score. RESULTS: A total of 123 patients who underwent isolated BT were included for analysis. The MCID, SCB, and PASS calculated for the American Shoulder and Elbow Surgeons score were 11.0, 16.8, and 59.6, respectively. For the Constant-Murley score, the calculated MCID and PASS were 3.8 and 19.5, respectively. The MCID, SCB, and PASS calculated for the SANE score were 3.5, 5.8, and 65.5, respectively. The following patient variables were significantly associated with decreased odds of achieving the MCID: workers' compensation status, male sex, and higher preoperative SANE score. Patients with a history of ipsilateral shoulder surgery had significantly reduced odds of achieving SCB. The only factor significantly associated with failing to reach the PASS was workers' compensation status. CONCLUSION: This study established values for the MCID, SCB, and PASS after BT without concomitant rotator cuff repair. Workers' compensation status, previous shoulder surgery, male sex, and higher preoperative patient-reported outcome measure scores are associated with lower odds of achieving clinically significant improvement after BT.

Preoperative Injections May Be an Iatrogenic Cause of Reoperation After Arthroscopic Rotator Cuff Repair.

PURPOSE: To determine whether an association exists between preoperative shoulder injections and reoperations in patients undergoing arthroscopic rotator cuff repair (aRCR). METHODS: The PearlDiver Patient Records Database was reviewed for Humana-insured patients undergoing aRCR after a shoulder injection. Two matched groups were created: aRCR within 1 year of injection (n = 12,054) and aRCR without prior injection within 1 year of surgery (n = 12,054). Reoperation rates within 3 months, at 3 to 6 months, and at 6 to 12 months postoperatively were assessed. Statistical analysis was performed with the χ-square test. RESULTS: The rate of reoperation within 3 months of the index procedure was higher in the control group (3.7% vs 3.1%, P = .01); however, 3 to 6 months after the index procedure, the rate of reoperation was higher in patients who received an injection within 1 year of the index procedure (1.8% vs 1.4%, P = .03). During the same intervals, the rate of revision rotator cuff repair (RCR) within 3 months of the index procedure was higher in the control group (2.9% vs 2.6%) and the rate of revision RCR 3 to 6 months after the index procedure was higher in patients who received an injection within 1 year of the index procedure (1.1% vs 0.9%); however, these results were not statistically significant (P = .3 and P = .8, respectively). The incidence of revision RCR (1.6% vs 1.1%; odds ratio, 1.4; P = .003) and incidence of subacromial decompression (1.5% vs 1.1%; odds ratio, 1.3; P = .01) 6 to 12 months after the index procedure were significantly higher in patients receiving an injection within 1 year before surgery. CONCLUSIONS: Preoperative shoulder injections may increase the risk of revision RCR and subacromial decompression by up to 150% in patients 6 to 12 months after index surgery compared with patients who did not receive a preoperative injection. However, the absolute increase in these revision procedures is only 0.5%. LEVEL OF EVIDENCE: Level III, comparative study.